Measurement of cortisol in fish scales is attracting considerable attention as a non-invasive indicator of chronic stress in wild populations. For many fish species of management and conservation interest, extensive scale collections exist that could provide extended records of individual stress responses, by combining cortisol measurements with life history information. However, it is not yet known how well cortisol is preserved in the scale during storage. To investigate the stability of scale cortisol, we accelerated potential degradation by storing scales from an individual farmed Atlantic salmon (Salmo salar) in an oven at 50°C for between 2 and 12 weeks. We found no significant relationship between scale cortisol concentration and either storage time or storage temperature. Cortisol concentrations in scales from the same fish were consistent (18.54–21.82 ng. g−1; coefficient of variation (CV) = 3.6%), indicating that scale cortisol can be reliably quantified, even in scales stored for varying periods of time or under different conditions. We also examined the effects of storage in real time using Atlantic salmon scales that were stored in paper envelopes at room temperature for between 3 and 32 years and found no significant relationship between scale cortisol concentration and storage time. Scale cortisol concentrations ranged from 4.05 to 135.37 ng.g−1 and levels of between-individual variability were high (CV = 61%). Given that scale cortisol does not degrade during long-term storage, historical scale collections and associated data describing fish life histories could potentially be used to develop bioindicators of physiological responses in fish populations. Further research is needed to understand scale cortisol variability and its biological relevance.